Heart failure (HF) is the leading cause of hospitalization in older adults. Multiple re-hospitalizations, many of which are preventable, are common in HF and contribute substantially to the cumulative burden of illness and cost of care. The rate of HF re-hospitalization is higher in African-American than in white patients, for reasons that are poorly understood. Current HF readmission risk prediction models are inadequate, and re-hospitalization prevention efforts are ineffective. Research is needed on novel risk factors for HF re-hospitalization. Depression is one of the best candidates for further investigation. It is a prevalent comorbidity in HF, and preliminary evidence suggests that it may increase the rate of re-hospitalization. In addition, the progression of HF and multiple re-hospitalizations may contribute to the persistence or worsening of depression. Thus, there may be a reciprocal relationship between depression and multiple re-hospitalizations. This study will be the first to investigate both sides of this relationship, and to examine the impact of depression on the long-term risk of multiple re-hospitalizations instead of focusing more narrowly on the short-term risk of the first re-hospitalization. The primary aim is to test the hypothesis that depression is an independent predictor of recurrent hospitalizations and mortality in patients with heart failure. The secondary aim is to test the hypothesis that multiple hospitalizations predict persistent or worsening depression in patients with HF. Additional aims include investigating whether depression helps to explain the disproportionately high rate of HF re-hospitalizations in African-American patients, and identifying mediators of the relationship between depression and re-hospitalization. A cohort of 400 adult patients (200 African-American, 200 white) with HF will be enrolled in the study and followed for 2 years after their index hospitalization at a large, urban teaching hospital in St. Louis. At enrollment, a structured interview to diagnose DSM-5 depressive disorders will be administered along with questionnaires to assess the severity of depression, anxiety, perceived stress, social support, HF self-care, and HF-related functional impairment. Data on prior HF hospitalizations, HF severity measures, comorbidities, and medications will be extracted from the electronic medical record (EMR) system. The depression questionnaire will be repeated every 3 months for 2 years. Hospitalizations and deaths will be ascertained by queries of the hospital's EMR system, interviews with the participants and collateral informants, medical records of admissions to other facilities, and the National Death Index. State-of-the-art methods for modeling recurrent events data will be used to analyze the time-dependent effect of depression on the rate of re-hospitalization and to determine whether it has more severe effects in African-American than in white patients. A linear mixed model will be used to evaluate the impact of multiple re-hospitalizations on the longitudinal course of depression. The findings will clarify the relationshp between depression and re-hospitalization in patients with heart failure, and help to identify novel targets for intervention.